Practical AI for engineers running real infrastructure
DevOps AI ToolKit exists because the gap between "AI hype" and "AI that actually helps you fix a stuck OpenStack VM at 2am" is real — and unnecessarily wide.
We publish prompt libraries, tool reviews, and DevOps guides aimed at one audience: the engineers actually running Linux, OpenStack, Kubernetes, GitLab CI, Prometheus, and the rest of the production stack. No fluff. No "10 AI tips that will change your life." Just the workflows that have worked in production.
James Joyner IV
Sr. Systems Software Engineer · San Jose, CA
I build and run large-scale, widely distributed Linux systems — Ubuntu or CentOS, and private clouds on OpenStack with Kolla-Ansible — and I live in the observability and on-call that comes with them (Prometheus and Grafana). I also support IaC with Terraform and long term storage such as Victoria Metrics. Gitlab for CI/CD. Hashicorp Vault for Security. Python for Administration. I started the DevOps AI ToolKit to share the AI workflows, prompts, and runbooks that actually survive contact with production.
What we cover
- AI prompt libraries tuned for infrastructure tasks (with safety constraints baked in).
- Honest, hands-on reviews of AI tools — Claude, ChatGPT, Cursor, Copilot, Warp, and more.
- Guides on combining AI with existing tools: Ansible, Terraform, GitLab CI, Prometheus, Grafana.
- Incident-response and SRE workflows that use AI without sacrificing safety.
What we don't do
- We don't recommend tools we haven't actually used.
- We don't pretend AI is magic. We treat it as a senior pair-programmer who occasionally hallucinates.
- We don't publish "AI will replace SREs" think-pieces. It won't. But it will change what good SREs do.
How we fund this
The guides, error references, validators, generators and the production-readiness auditor are free and stay free. The work is funded by four things that already exist, not by anything planned:
- Pro membership — unlimited AI incident triage, code review, monitoring generation, and saved troubleshooting history.
- Prompt packs and kits — one-time downloadable packs, cheat sheets and runbook bundles.
- Fixed-price audits and consulting, plus managed OpenStack operations and team training.
- Affiliate links on tool reviews — we only recommend tools we'd run on production infrastructure ourselves.
See our affiliate disclosure for full details on the affiliate relationships, and pricing for what each paid tier includes.
Get in touch
Have a workflow that worked? Want to suggest a prompt? Disagree with a review? Reach out — we read everything.
Evaluating the platform for your team?
Security practices, live system status, support levels, the product changelog, and enterprise capabilities — with existing features clearly separated from planned ones.